A person’s social network constitutes a rich sampling space for informing judgments about social statistics (e.g., the distribution of preferences, risks, or behaviors in the broader social environment). How is this sampling space searched and used to make inferences? This chapter gives an overview on research on the social-circle model, a computational process account of how people make inferences about relative event frequencies. The social-circle model is inspired by the notion of sequential and limited search in models of bounded rationality for multi-attribute decision making. In accord with research on the structure of social memory, the model assumes that social sampling proceeds by sequentially probing a person’s social circles – including oneself, family, friends, and acquaintances – and that search is constrained by a simple stopping rule. The social-circle model has several free parameters that enable it to capture individual differences in the order in which social circles are inspected, in noise during evidence evaluation, and in discrimination thresholds. We provide a step-by-step tutorial for deriving predictions of the social-circle model, review empirical tests of the model, illustrate how the model reflects individual differences in social sampling and differences in sampling across domains, and analyze the ecological rationality of heuristic social sampling.